Approximately 68% of adults and 34% of children in the United States are overweight or obese with an attendant increased risk of adiposity-related co-morbidities. The salient phenotypes (primarily increased energy intake) that promote weight gain are more evident during- than after- weight gain. Therefore, analyses of the mechanisms leading to weight gain are more likely to be illuminating in studies of children who are at high vs. low risk of becoming obese adults as opposed to studies of adults who are already overweight compared to lean subjects. An extremely common (allelic frequency of ~40%) SNP at the FTO locus (rs9939609) that is associated with early onset obesity and increased food intake in experimental settings has recently been identified. In addition, recent studies document that: 1) Changes in eating behavior can be detected in structured, objective meal paradigms among children with allelic variants that place them at risk for obesity, but who are not yet obese. The central hypotheses of these studies are: 1) Energy intake and mean energy density of selected foods are increased in children with one (AT) or two (AA) copies of the rs9939609 obesity- risk (A) SNP of FTO/FTM compared to those who are unaffected (TT), 2) Using functional magnetic resonance imaging (fMRI), affected children (AA or AT) will demonstrate enhanced ventral front striatal responses to food stimuli and/or reduced inhibitory input from prefrontal centers compared to those without it, showing a dose response of the at-risk allele (AA>AT>TT). Behavioral and neuroimaging differences by genotype identified at baseline will be correlated with subsequent weight gain during a 2-year longitudinal assessment period. Over 5 years we plan to study and follow 30 trios consisting of gender-, age-, and weight-matched 6-9 year olds (1 AA, 1 AT, and 1 TT). These studies are unique in examining eating behavior and the functioning of relevant CNS circuits among individuals at varying degrees of risk for obesity who, at the time of study, will be pre- obese, thus avoiding the confounds of obesity per se on these behavioral and imaging phenotypes.